Re: [PATCH v3 0/3] automatically skip away from broken commits

[Christian Couder <christian.couder@gmail.com>]

On Tue, Jun 9, 2009 at 10:37 PM, Junio C Hamano<gitster@pobox.com> wrote:
> Christian Couder <christian.couder@gmail.com> writes:
>
>> My opinion is that we should not penalize all the people working on
>> "quite clean" projects and also people working on "not clean" projects
>> who are able to recover, on the pretence that there are other people
>> on these "not clean" projects who are not.
>>
>> I think it's the projects maintainers' responsibility to keep their
>> projects graphs quite clean (and they have the right to ask git
>> developers for the tools to do that).
>
> You seem to be saying that a completely linear history is the only one
> that is clean.

I was thinking about projects that use test suites to make sure new
commits don't break a lot of things.
But I should have said "to keep their projects graph quite clean or to
provide specific information about how to work around potential
problems when bisecting."

> In an earlier message, I illustrated a case where two people independently
> worked on unrelated topic and ended up merging their branches together, to
> illustrate why your "not adjacent in goodness space" algorithm does not
> give "away from untestable ones in topology space".
>
> With your definition, that history is _not_ clean.  I do not think any
> project wnats that kind of cleanness in their history.  You just made the
> word "clean" to describe the history meaningless.

When I wrote "clean", I just mean with not too many untestable commits.

> My take on the issue you mentioned above is that we should not penalize
> the codepath's simplicity too much, only to cater to pathological
> behaviour exhibited on an uninteresting special case of competely linear
> history.

I tried to make a short yet efficient implementation, not to favor a
special case.

> Your algorithm is Ok as an initial cut, in that it is an improvement over
> the "next in goodness space, not even bother to avoid the pathological
> case" algorithm.  But I do not think it is much better than HPA's "try the
> best one if it is not skipped, otherwise pick one randomly", and if we
> wanted to do better and to claim that we pick ones that are _away_ from
> untestable ones, we do need to take topology into account.  That is all I
> am saying.

Ok. I started working on optionaly using a PRNG but I am not sure that
you will want to add another one.

Best regards,
Christian.